V2 DRM A/B testing measures whether DRM affects sales/piracy rates with protected vs DRM-free cohorts. The infrastructure uses aggregate storefront sales metrics and privacy-safe piracy-rate proxy metrics; it does not export raw account identifiers or automate penalties.
sales/piracy rates. aggregate storefront sales metrics. no raw PII export.
Experiment Coverage#
V2/ue/Build/DRM/v2-drm-ab-testing.json defines the approved DRM A/B testing
experiments. Each experiment has a ProtectedDRM variant and a DRMFree variant, a
storefront-region-day assignment unit, a 28-day analysis window, a minimum
sample size of 10000, and 95 percent confidence reporting.
Sales impact uses aggregate unit sales, revenue per visitor, and conversion
rate. Piracy-rate impact uses aggregate anti-tamper telemetry, activation-gap,
and unauthorized-launch proxy metrics. The DRM-free variants depend on
V2/ue/Build/DRM/v2-drm-removal-pathway.json, and every experiment must account
for the performance confound checks from
V2/ue/Build/DRM/v2-drm-impact-benchmarking.json.
Privacy And Approval#
DRM A/B tests require pre-registration, privacy-preserving aggregation, no raw PII export, legal privacy review, publishing approval, security review, human review, public result summaries, and a rollback runbook. Anti-tamper telemetry is used only as an aggregate piracy-rate proxy and remains subject to manual review before any penalty decision.
CI Gates#
V2/ue/Tools/plan-v2-drm-ab-testing.py emits the deterministic DRM A/B testing
plan. V2/ue/Tools/check-v2-drm-ab-testing.py validates the manifest, planner
output, DRM removal dependency, impact benchmarking dependency, anti-tamper
telemetry dependency, privacy guardrails, Horde wiring, CI workflow wiring, and
this runbook.
Required release gates:
drm-ab-testing-plan-generateddrm-ab-testing-experiment-coveragedrm-ab-testing-variant-coveragedrm-ab-testing-sales-metric-coveragedrm-ab-testing-piracy-metric-coveragedrm-ab-testing-privacy-guardrails